The proposed research will examine episodes of mindless reading, with special focus on the novel insights that such episodes may offer into bilingual cognition. Mindless reading (MR) is a daydream mode wherein readers continue moving their eyes across the text, but their attention has disengaged from text processing. In some regards, MR is similar to other automatized processes, such as driving a familiar route while having a conversation. People are adept at decoupling thought from action, allowing well-practiced behaviors to proceed in the background, while the conscious mind is elsewhere occupied. Indeed, it is estimated that mind-wandering occupies anywhere from 30-50% of everyday cognition. In recent years, numerous studies have documented a broad set of cognitive benefits that occur in bilingualism, ranging from superior low-level, perceptual categorization to superior high-level executive functioning. A prominent hypothesis is that bilinguals constantly engage neural circuitry to control language selection, even when the dominant language is being used. This juggling act promotes robust neural systems for cognitive control, even in non-linguistic tasks. The proposed research is a synthesis of these fields, and represents the first investigation of mind-wandering (or MR) in bilinguals. We will conduct foundational experiments on MR in monolinguals and bilinguals, as an ecologically valid means to investigate one potential source of their cognitive differences. According to recent theory, mind-wandering reflects activity of the brain's default network, which cyclically trades off with more controlled activity in frontal-parietal networks. We suggest a working hypothesis that, as a function of constant language selection, bilinguals are less likely to mind-wander, especially during language processing. We will also test the counter-intuitive hypothesis that mind-wandering will occur more often when people read in their dominant language. The scientific study of mind-wandering poses methodological challenges, as it often requires self-report by the participant, who is typically unaware that attention has lapsed. In MR, however, there are characteristic changes in eye-movements that allow early, covert detection. Our research will use eye-tracking to identify MR in an unobtrusive manner during natural reading. Our specific aims are (1) to continue developing a Bayesian network that classifies eye-movement patterns as indicating MR, in real-time, without requiring participants to self-report MR or respond to external prompts; (2) to determine the frequency and duration of MR episodes during natural reading, with special focus on bilinguals reading in their dominant and non-dominant languages; (3) to determine whether monolinguals and bilinguals differ in detection rates for gaze-contingent display changes during MR; (4) to directly assess the role of cognitive load in MR, by comparing MR rates when passages are presented in print versus natural handwriting, which increases the need for top-down processing in reading; and (5) to complete theoretical simulations of the experiments using the E-Z Reader model of eye-movement control.